The layered manufacturing (LM) based locking bone plates (LBPs) are highly favored for biomedical applications due to its implant customization flexibility, control over porosity, and functional parameters. Design of experiments mainly relies on predefined experimental run orders; however, machine learning (ML) can handle complex and nonlinear relationships for predicting output responses. This research investigates the application of various ML models for predicting the impact strength, torque values, and punch shear strength of poly lactic acid (PLA) based LBPs. A dataset of 100 data points for each response variable was developed, analyzing the influence of printing parameters, like infill density (ID), layer height (LH), wall thickness (WT), and print speed (PS). The ID, LH, WT, and PS were varied within the ranges of 20%–100%, 0.1–0.5 mm, 0.4–1.2 mm, and 20–100 mm/s, respectively. Among the ML models evaluated, XGBoost and Adaboost exhibited strong alignment of the predicted values with actual values. The decision tree regression, showed the lowest predictive accuracy due to its tendency to overfit and lack of iterative refinement. The findings suggest that ML models can effectively predict the mechanical properties of PLA‐based LBPs, offering valuable insights for optimizing printing parameters to enhance the performance of orthopedic implants. The findings can guide biomedical engineers in making data‐driven decisions to enhance the strength of LBPs, thereby improving patient outcomes in orthopedic treatments.Highlights
Predicted impact strength, torque, and shear strength of biomedical specimens.
Evaluated 100 data points per variable to assess printing parameters' influence.
XGBoost and Adaboost showed superior accuracy, with R2 consistently above 0.9.
Decision Tree regression exhibited the lowest accuracy due to overfitting issues.
Aims to guide biomedical engineers in data‐driven decisions for better patient outcomes.